132 Prof. \V. Thomson on the Reduction of 



from the true average value 



of a / —r for the portion of rock 



between the extreme thermometers. 



36. Dividing tt by the square of the reciprocal of this number, 



k 

 we find 235 '1 as the value of -, or, as we may call it, the con- 

 ductivity of the rock in terms of the thermal capacity of a cubic 

 foot of its own substance. In other words, we infer that all the heat 

 conducted in a year (the unit of time) across each square foot of 

 a plate one French foot thick, with its two sides maintained con- 

 stantly at temperatures differing by 1°, would, if applied to raise 

 the temperature of portions of the rock itself, produce a rise of 

 1° in 235 cubic feet. As it is difficult (although by no means 

 impossible) to imagine circumstances in which the heat, regu- 

 larly conducted through a stratum maintained, with its two sides, 

 at perfectly constant temperatures, could be applied to raise the 

 temperatures of other portions of the same substance, we may 

 vary the statement of the preceding result, and obtain the fol- 

 lowing completely realizable illustration. 



37. Let a large plate of the rock, everywhere one French foot 

 thick, have every part of one of its sides (which, to avoid circum- 

 locution, we shall call its lower side) maintained at one constant 

 temperature, and let portions of homogeneous substance, at a 

 temperature 1° lower, be continually placed in contact with the 

 upper surface, and removed to be replaced by other homogeneous 

 portions at the same lower temperature, as soon as the tempera- 

 ture of the matter actually thus applied rises in temperature by 

 ToVu °f a degree. If this process is continued for a year, the 

 whole quantity of the refrigerating matter thus used to carry 

 away the heat conducted through the stratum must amount to 

 235,000 cubic feet for each square foot of area, which will be at 

 the rate of -00745 of a cubic foot per second. We may there- 

 fore imagine the process as effected by applying an extra stratum 

 •00745 of a foot thick every second of time. This extra stratum, 

 after lying in contact for one second, will have risen in tempera- 

 ture by y^jjtf of a degree. By means of the information con- 

 tained in this apparently unpractical statement, many interesting 

 problems may be practically solved, as I hope to show in a sub- 

 sequent communication. 



38. The value of a / — , derived from the whole eighteen 



years' period of observation (-1150), differs so little from that 

 (•1154) found previously (§ 16) from Professor Forbes's observa- 

 tions and reductions of the first five of the years, that we may 

 feel much confidence in the accuracy of the values '1098 and 



